Author + information
- Received October 23, 2014
- Revision received March 25, 2015
- Accepted April 20, 2015
- Published online June 30, 2015.
- Satoru Kishi, MD∗,
- Gisela Teixido-Tura, MD, PhD†,
- Hongyan Ning, MD‡,
- Bharath Ambale Venkatesh, PhD∗,
- Colin Wu, PhD§,
- Andre Almeida, MD∗,
- Eui-Young Choi, MD∗,
- Ola Gjesdal, MD∗,
- David R. Jacobs Jr., PhD‖,
- Pamela J. Schreiner, PhD‖,
- Samuel S. Gidding, MD¶,
- Kiang Liu, PhD‡ and
- João A.C. Lima, MD∗∗ ()
- ∗Division of Cardiology, Johns Hopkins University, Baltimore, Maryland
- †Department of Cardiology, Hospital Vall d’Hebron, Barcelona, Spain
- ‡Department of Preventive Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois
- §Office of Biostatistics Research, National Heart, Lung, and Blood Institute, Bethesda, Maryland
- ‖Division of Epidemiology and Community Health, School of Public Health, University of Minnesota School of Public Health, Minneapolis, Minnesota
- ¶Nemours Cardiac Center, Alfred I. duPont Hospital for Children, Wilmington, Delaware
- ↵∗Reprint requests and correspondence:
Dr. João A.C. Lima, Johns Hopkins University School of Medicine, Division of Cardiology, 600 North Wolfe Street, Blalock 524, Baltimore, Maryland 21287-8222.
Background Cumulative blood pressure (BP) exposure may adversely influence myocardial function, predisposing individuals to heart failure later in life.
Objectives This study sought to investigate how cumulative exposure to higher BP influences left ventricular (LV) function during young to middle adulthood.
Methods The CARDIA (Coronary Artery Risk Development in Young Adults) study prospectively enrolled 5,115 healthy African Americans and whites in 1985 and 1986 (baseline). At the year 25 examination, LV function was measured by 2-dimensional echocardiography; cardiac deformation was assessed in detail by speckle-tracking echocardiography. We used cumulative exposure of BP through baseline and up to the year 25 examination (millimeters of mercury × year) to represent long-term exposure to BP levels. Linear regression and logistic regression were used to quantify the association of BP measured repeatedly through early adulthood (18 to 30 years of age) up to middle age (43 to 55 years).
Results Among 2,479 participants, cumulative BP measures were not related to LV ejection fraction; however, high cumulative exposure to systolic blood pressure (SBP) and diastolic blood pressure (DBP) were associated with lower longitudinal strain rate (both p < 0.001). For diastolic function, higher cumulative exposures to SBP and DBP were associated with low early diastolic longitudinal peak strain rate. Of note, higher DBP (per SD increment) had a stronger association with diastolic dysfunction compared with SBP.
Conclusions Higher cumulative exposure to BP over 25 years from young adulthood to middle age is associated with incipient LV systolic and diastolic dysfunction in middle age.
Blood pressure (BP) at the higher end of the population distribution may represent a chronic exposure that injures the cardiovascular system. Cumulative BP exposure from young adulthood to middle age may adversely influence myocardial function and predispose individuals to heart failure (HF) and other cardiovascular disease later in life. The 2005 guidelines for the diagnosis and treatment of HF from the American College of Cardiology and American Heart Association highlighted the importance of early recognition of subclinical cardiac disease and of noninvasive tests in the clinical evaluation of HF (1).
Speckle-tracking echocardiography (STE) offers the opportunity to evaluate subclinical markers of left ventricular (LV) dysfunction, and global myocardial strain assessed by STE has been described as a better predictor of future HF and other cardiovascular endpoints compared with left ventricular ejection fraction (LVEF) (2,3). Hypertension has been related to reduced global myocardial deformation as a precursor of HF (4,5), and, in turn, global myocardial deformation parameters (longitudinal and circumferential shortening and radial thickening) have been directly related to HF (6,7).
Our main objective was to investigate how cumulative exposure to high BP from young to middle adulthood influences LV function. In the CARDIA (Coronary Artery Risk Development in Young Adults) study, multiple repeated measures of BP and other cardiovascular risk factors were recorded over a 25-year timespan, starting during early adulthood (ages 18 to 30 years). In this report, we demonstrate the cumulative effect of BP exposure to conventional echocardiographic measures of LV systolic and diastolic function, as well as LV strain measures assessed by STE, also used as markers of subclinical LV dysfunction.
The CARDIA study is a multicenter prospective investigation that enrolled 5,115 healthy African-American and white men and women in 1985 and 1986 (year 0) from 4 U.S. field centers (Birmingham, Alabama; Oakland, California; Chicago, Illinois; Minneapolis, Minnesota). Participants were followed prospectively in 7 subsequent examinations. The Institutional Review Boards at each study site approved the study protocols, and written informed consent was obtained from all participants. Of 3,498 participants attending the year 25 (2010 to 2011) examination, 3,474 underwent standard echocardiography and STE assessment. We excluded participants with incomplete BP measurement during 25 years of follow-up (n = 995). The remaining 2,479 participants with complete longitudinal datasets were included in the analyses. At baseline, the included participants were slightly older (25.3 vs. 24.5 years) and more female than male (1,409 vs. 1,070), and had a lower percentage of blacks (41.4% vs. 61.1%), higher educational attainment (14.2 vs. 13.4 years), and lower baseline systolic blood pressure (SBP) (109.8 vs. 110.9 mm Hg; all p < 0.01) compared with those not included in this study (n = 2,636) (Online Table 1).
Standardized protocols were used to measure height, weight, heart rate, BP, lipids, glucose, smoking status, educational level, and physical activity (8,9). Weight and height were measured using a standard balance beam scale. Body mass index was calculated as weight (kilograms) divided by height in meters squared. We used the average of the second and third of 3 BP measurements using random-zero sphygmomanometry performed at 1-min intervals after the participant had been sitting quietly for 5 min in a still room (year 0 to 15) and a BP monitor (year 20 to 25; Omron Healthcare Inc., Lake Forest, Illinois) (10). A calibration study was performed, and values calibrated to the sphygmomanometric measures were used for year 20 and year 25 BP measurements. We used recalibrated measures (10). Pulse pressure (PP) as the index of pulsatile component of BP was calculated as SBP minus diastolic blood pressure (DBP); mean arterial pressure (MAP) was calculated as: DBP + (1/3 PP). The presence of diabetes was assessed at each examination on the basis of a combination of medication use for diabetes (every examination), fasting plasma glucose ≥126 mg/dl (examination years 0, 7, 10, 15, 20, and 25), 2-h glucose ≥200 mg/dl (years 10, 20, and 25), or glycosylated hemoglobin ≥6.5% (years 20 and 25). Total cholesterol was measured in fasting plasma samples and determined by enzymatic procedures.
We calculated the cumulative exposure of BP for each participant over 25 years from age 18 to 30 years to age 43 to 55 years as millimeters of mercury × year at each clinic visit to represent long-term exposure to BP levels. We defined this BP product for each participant’s accumulated exposure to BP. We used the area under the BP curve over 25 years as a covariate in the univariable and multivariable analyses. Calculations were performed for SBP, DBP, PP, and MAP.
Echocardiographic and speckle-tracking analysis
Experienced sonographers performed Doppler echocardiography and 2-dimensional–guided M-mode echocardiography using an Artida cardiac ultrasound scanner (Toshiba Medical Systems, Otawara, Japan) using standardized protocols across all field centers. The sonographers made measurements from digitized images using a standard off-line image analysis software system (Digisonics, Inc., Houston, Texas). LVEF and left atrial volume were measured from the apical 4-chamber view on the basis of the American Society of Echocardiography guidelines; LV mass was derived from the Devereux formula using the American Society of Echocardiography guidelines (11). Left atrial volume and LV mass were indexed to body surface area (left atrial volume index and left ventricular mass index [LVMI]). In the diastolic functional parameters, peak velocities of the early phase (E), late phase (A) of the mitral inflow, and their ratio (E/A ratio) were measured from pulsed Doppler echocardiography recordings of transmitral flow; early peak diastolic mitral annular velocity (e’) at the septal mitral annulus was measured using tissue Doppler imaging (12). E/e’ ratio was calculated as an index of LV filling pressures.
STE for myocardial strain and strain rate measurements was performed offline with dedicated semiautomated 2-dimensional STE software (Toshiba Medical Systems) only at year 25 (13). Three cardiac cycles from each view were recorded for offline analyses. Average frame rate was 46.2 frames per second with mean heart rate <70 beats/min on the basis of consensus (14). Strain was calculated as the peak systolic change in segment length relative to its end-diastolic value. Longitudinal strain and strain rate curves were assessed from 4-chamber views. Circumferential strain and strain rate were assessed from the short-axis view at mid-ventricular level. STE images were analyzed on a 6-segment basis for endocardial, mid-wall, and epicardial contours for each view. The mid-wall layer in each segment was analyzed for global longitudinal and circumferential strain and strain rate. The image set in each view was considered inadequate for STE if 4 or more segments had been excluded because of less optimal tracking. Global strain values were calculated as the average of segmental peak strains. Global strain rate values were also calculated from the average of segmental peak values for each phase (in s−1). STE indices of systolic cardiac deformation at the year 25 examination included 4-chamber longitudinal peak systolic strain (Ell), circumferential peak systolic strain (Ecc), radial peak systolic strain (Err), 4-chamber longitudinal systolic strain rate (Ell_SRs), circumferential systolic strain rate (Ecc_SRs), and radial systolic strain rate (Err_SRs). Diastolic STE indices were peak early diastolic strain rate in the 4-chamber longitudinal (Ell_SRe), circumferential (Ecc_SRe), and radial (Err_SRe).
Descriptive statistics were displayed using means and standard deviations for continuous variables and group proportions for categorical variables. Student t tests were used to compare means between blacks and whites of both sexes. Multivariable linear regression models assessed the longitudinal associations between year 0 BP parameters and STE parameters and conventional echocardiographic parameters measured at the year 25 examination. Multivariable models were adjusted for the following traditional cardiovascular disease risk factors: Model 1 included age at year 25, sex, race, and field centers as demographic variables; Model 2 included Model 1 and average level of body mass index (kilograms/meters squared), physical activity (exercise unit), alcohol intake (milliliters), heart rate (beats/min), total cholesterol (milligrams/deciliter), diabetes, and current smoking status at year 25 as cardiovascular risk factors; Model 3 included Model 2 and LVMI as an LV structural parameter; and Model 4 included Model 3 and use of antihypertensive medication at year 25.
For categorical analyses of clinically significant cardiac dysfunction, systolic dysfunction was defined as LVEF <50% and diastolic dysfunction as E/e’ ratio >15 (15–17). We explored relationships between BP (per SD higher) over 25 years and clinically relevant LV dysfunction at year 25 using multivariable logistic regression analysis, which we evaluated accounting for similar covariates, reporting odds ratios (ORs) and 95% confidence intervals (CIs).
We explored the relationship between the deciles of cumulative BP exposure and standardized LV function. We compared the deciles of cumulative BP for LV function compared with the lowest BP decile (0% to 10%) group.
A 2-sided p value of <0.05 was considered for statistical significance. All statistical analyses were performed using SAS version 9.3 for Windows (SAS Institute Inc., Cary, North Carolina).
Cohort characteristics at year 25 are shown in Table 1. The mean age at this time point was 50 ± 3.6 years, 56.8% of all participants were women, and 60.0% were white. Mean values for conventional echocardiographic parameters were within the normal range in mid-age (43 to 55 years). In this CARDIA cohort, SBP increased gradually after 35 years of age and DBP and MAP increased from baseline to the fourth decade and then plateaued (Central Illustration, Online Figure 1). PP decreased until 40 years of age and then increased after age 40 years. Black men and women had higher SBP, DBP, and MAP levels compared with whites more than 25 years of age. Men had higher PP than women.
LV systolic and diastolic function
LV systolic dysfunction (LVEF <50%) was present in 70 participants, diastolic dysfunction (E/e’ ratio >15) was present in 69 participants, and systolic or diastolic dysfunction was present in 135 participants. Cumulative BPs (SBP, DBP, MAP, and PP) were not associated with the likelihood of having an ejection fraction <50% measured by conventional echocardiography (Table 2, Online Table 2); meanwhile, higher SBP (OR: 1.46; 95% CI: 1.09 to 1.94), DBP (OR: 1.69; 95% CI: 1.23 to 2.33), and MAP (OR: 1.40; 95% CI: 1.10 to 1.78) were predictors of diastolic dysfunction after full adjustment. Higher DBP had the strongest association with diastolic dysfunction. Both higher SBP (OR: 1.38; 95% CI: 1.10 to 1.73) and DBP (OR: 1.34; 95% CI: 1.05 to 1.70) were predictors of the presence of systolic or diastolic dysfunction after full adjustment.
Higher deciles of cumulative SBP and DBP were associated with low Ell_SRs (p < 0.05) compared with the lowest BP decile; however, neither SBP nor DBP was associated with LVEF (Central Illustration). Results for MAP were similar (Online Figure 2). Cumulative exposure to BP parameters (SBP, DBP, and PP) was not related to LVEF measured by conventional echocardiography (Table 3, Online Table 3); however, high cumulative exposure to SBP, DBP, and MAP were independently associated with low Ell and Ell_SRs. The magnitudes of the standardized β-coefficients for DBP (0.119 for Ell and 0.107 for Ell_SRs; all p < 0.001) in relation to Ell and Ell_SRs were higher than those for SBP (0.078 for Ell; p < 0.05; and 0.086 for Ell_SRs; p < 0.005) and MAP (0.112 for Ell; p < 0.001; and 0.080 for Ell_SRs; p < 0.005). High cumulative exposure to PP was independently associated with Ecc (standardized β-coefficient = −0.06; p < 0.05). High cumulative exposure to SBP (0.118 for Err and 0.115 for Err_SRs; all p < 0.001) and PP (0.105 for Err and 0.106 for Err_SRs; all p < 0.001) was independently associated with high Err and Err_SRs. LVMI attenuated the relationship of BP to LV systolic function in Model 3, but antihypertensive medications did not attenuate these relationships in Model 4. Prediction of LV function from BP at year 0 (baseline) was less powerful than cumulative BP over 25 years (Online Table 4).
Higher deciles of cumulative SBP and DBP were associated with increased E/e’ ratio (p < 0.05), and higher deciles of cumulative SBP and DBP were associated with reduced Ell_SRe (p < 0.05) compared with the lowest BP decile (Central Illustration). Similar relationships were found for MAP (Online Figure 2). High cumulative exposure to SBP, DBP, and MAP was independently associated with low E/A ratio (SBP: −0.157, DBP: −0.217, MAP: −0.212; all p < 0.001), Ell_SRe (SBP: −0.151, DBP: −0.181, MAP: −0.185; all p < 0.001), and high E/e’ ratio (SBP: 0.183, DBP: 0.161, MAP: 0.186; all p < 0.001) (Table 4, Online Table 5). High cumulative exposure to SBP and PP was independently associated with high left atrial volume index (SBP: 0.068; p < 0.05; PP: −0.127; p < 0.001). LVMI attenuated the relationship of BP to LV diastolic function parameters in Model 3, but use of antihypertensive medications did not (Model 4).
This study demonstrates the effects of chronically elevated BP—expressed as cumulative exposure to hypertension from young adulthood to middle age—on LV function measured at middle age. The study documents a greater effect of cumulative BP exposure on diastolic versus systolic myocardial function in a population of healthy adults, with a greater effect of diastolic as opposed to SBP elevation on both systolic and diastolic LV dysfunction.
Cumulative effects of BP on myocardial deformation
LV function has been reported to be associated with the effects of hypertension and other chronic factors such as obesity (13,18–21). In our study, cumulative BP was not associated with reduced LVEF (<50%); however, cumulative exposure to elevated BP, especially high DBP and MAP, was independently associated with reduced global systolic myocardial LV function as assessed by Ell and Ell_SRs. In addition, Ell_SRe was affected by mildly elevated cumulative BP compared with E/e’ ratio, which was affected by high cumulative BP exposure (Central Illustration). Thus, Ell_SRe may be a good predictor of diastolic LV dysfunction even when compared with E/e’ ratio.
In an older population, the MESA (Multi-Ethnic Study of Atherosclerosis) demonstrated a relation between elevated DBP and decreased regional myocardial function using cardiac magnetic resonance (5). Likewise, high DBP at an older age was associated with reduced longitudinal myocardial shortening as measured by STE in the Framingham Study (22). In hypertensive patients with LV hypertrophy and normal ejection fraction, Palmon et al. (23) reported reduced intramural (endocardial) longitudinal and circumferential shortening. Worse longitudinal and circumferential myocardial deformation also have been associated with adverse outcomes (6).
Echocardiographic measurement of worsening of longitudinal and circumferential strain mostly represents reduced longitudinal and circumferential deformation of endocardial fibers, respectively. This may reflect a chronic process that begins with cumulative BP effects increasing afterload and aortic stiffness, thus increasing systolic LV wall stress and inducing LV remodeling (4,24). Compensatory mechanisms activated to maintain cardiac contraction include: 1) an increase in LV mass with concentric remodeling (25); and 2) an increase in LV torsion (4). Increased LV wall stress damages the oblique fiber structure in the endocardium; this may lead to myocardial fibrosis (26). Moreover, long-term exposure to high BP stimulates reactive interstitial fibrosis, as shown in hypertensive cardiomyopathy (25,27). Reduced subendocardial shortening, observed in this study using STE, is the first sign of impairment of longitudinal myocardial function. Compensation by mid-wall fibers with resultant improvement in radial contractility maintains overall cardiac function and explains the lack of relation to conventional systolic function parameters (28).
In young adulthood (before 40 years of age), aging is associated with an increase in DBP and a decrease in PP up to age 40 years (Central Illustration, Online Figure 1). After age 40 years, aging is associated with a decrease in DBP and an increase in PP; in the latter half of the human life span SBP increases with each decade of life (29,30). Of note, our findings show a turning point for PP that occurs earlier than that demonstrated in the Framingham Study (30). Furthermore, BP, especially elevated DBP, already produced worsening myocardial function in healthy young adulthood. An elevation in DBP may be useful as an outcome predictor in young adulthood in contrast to the greater importance of SBP in older individuals (31).
This study shows that the use of speckle-tracking data and LV dysfunction help us better understand the adverse effects of cumulated BP exposure that ejection fraction alone does not identify.
In prior CARDIA reports, Pletcher et al. (32) and Allen et al. (33) demonstrated that a higher BP trajectory beginning in young adulthood increased the risk of the prevalence of coronary calcium in middle age. These studies suggest that long-term BP control in early young adulthood may be important to prevent coronary heart disease. Our findings suggest that the relationship of long-term exposure to higher BP and LV dysfunction is similar to that for subclinical atherosclerosis and provides further support for the importance of good risk factor control early in life. In the CARDIA study, potentially modifiable antecedents of HF were present in African Americans more than a decade before the onset of clinical HF (31). Each increase of 10 mm Hg in DBP among African Americans 20 to 29 years of age had a 2-fold greater risk for development of HF at 40 to 49 years of age.
Long-term treatment of both systolic and diastolic hypertension reduces the risk of HF (34,35). Of patients with hypertension, 20% to 30% are unaware of their condition (36,37). In addition, 50% to 70% of all patients with hypertension do not achieve their target BP (36–38). In the National Longitudinal Study of Adolescent Health, Nguyen et al. (39) reported that approximately 1 in 5 young adults had high BP. In our cohort, many participants were not hypertensive at age 18 to 30 years; however, chronic exposure to higher BP, even within what is considered the normal range, is independently related to LV dysfunction 25 years later. Young adults may prevent the development of LV dysfunction by reducing sodium intake, keeping ideal body weight, and performing adequate levels of physical activity, as well as by adhering to medical treatment (40–42). Clinical trials of earlier treatment for hypertension may be warranted to determine whether favorable cardiac function can be maintained by decreasing cumulative BP exposure. Our findings and those of some previous studies suggest that prehypertension throughout early adulthood is harmful to cardiac function (4,20,22).
We relied on a subclinical endpoint (LV systolic and diastolic dysfunction measured by echocardiography); our findings will need validation using clinical endpoints. We did not aim to expose the potential effect of differences between classes of antihypertensive medication on LV function. Our participants were 18 to 30 years of age at baseline (year 0); thus, our study could not assess cumulative BP exposure at earlier ages. This study was not designed to identify a specific BP early in age that was associated with future myocardial dysfunction. Also, all analyses were performed on a single system (Toshiba Medical Systems). Recently, the inter/intrasonographer reproducibility for STE analysis using the Artida cardiac ultrasound scanner in the CARDIA Study has been published (43). The results were highly reproducible in this large cohort study. However, recent publications have shown significant intervendor variability of strain (14).
Cumulative exposure to BP over 25 years from young adulthood to middle age is associated with incipient LV systolic and diastolic dysfunction in middle age. In early adulthood, DBP control may be of particular importance to prevent LV dysfunction and HF.
COMPETENCY IN MEDICAL KNOWLEDGE: BP in early adulthood is related to later systolic and diastolic ventricular dysfunction.
TRANSLATIONAL OUTLOOK: Further studies are needed to determine whether specific lifestyle interventions implemented to reduce DBP in early adulthood prevent HF from developing later in life.
The authors thank the other investigators, staff, and participants of the CARDIA study for their valuable contributions.
The CARDIA Study is supported by Contracts HHSN268201300025C, HHSN268201300026C, HHSN268201300027C, HHSN268201300028C, HHSN268201300029C, and HHSN268200900041C from the National Heart, Lung, and Blood Institute, the Intramural Research Program of the National Institute on Aging, and an intra-agency agreement between the National Institute on Aging and the National Heart, Lung, and Blood Institute (AG0005). The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- blood pressure
- confidence interval
- diastolic blood pressure
- circumferential peak systolic strain
- circumferential peak early diastolic strain rate
- circumferential peak systolic strain rate
- 4-chamber longitudinal peak strain
- 4-chamber longitudinal peak early diastolic strain rate
- 4-chamber longitudinal peak systolic strain rate
- radial peak strain
- radial peak systolic strain rate
- heart failure
- left ventricular
- left ventricular ejection fraction
- left ventricular mass index
- mean arterial pressure
- odds ratio
- pulse pressure
- systolic blood pressure
- speckle-tracking echocardiography
- Received October 23, 2014.
- Revision received March 25, 2015.
- Accepted April 20, 2015.
- American College of Cardiology Foundation
- Hunt S.A.
- Wong C.Y.,
- O’Moore-Sullivan T.,
- Leano R.,
- Byrne N.,
- Beller E.,
- Marwick T.H.
- Yoneyama K.,
- Gjesdal O.,
- Choi E.Y.,
- et al.
- Rosen B.D.,
- Saad M.F.,
- Shea S.,
- et al.
- Ersbøll M.,
- Valeur N.,
- Mogensen U.M.,
- et al.
- Choi E.Y.,
- Rosen B.D.,
- Fernandes V.R.,
- et al.
- Lang R.M.,
- Bierig M.,
- Devereux R.B.,
- et al.
- Kishi S.,
- Armstrong A.C.,
- Gidding S.S.,
- et al.
- Goldberg L.R.,
- Jessup M.
- From A.M.,
- Scott C.G.,
- Chen H.H.
- Huffman M.D.,
- Berry J.D.,
- Ning H.,
- et al.
- Lloyd-Jones D.M.,
- Larson M.G.,
- Leip E.P.,
- et al.
- Kishi S.,
- Reis J.P.,
- Venkatesh B.A.,
- et al.
- Cheng S.,
- Larson M.G.,
- McCabe E.L.,
- et al.
- Palmon L.C.,
- Reichek N.,
- Yeon S.B.,
- et al.
- Rosen B.D.,
- Edvardsen T.,
- Lai S.,
- et al.
- Mewton N.,
- Liu C.Y.,
- Croisille P.,
- Bluemke D.,
- Lima J.A.
- Jellis C.,
- Martin J.,
- Narula J.,
- Marwick T.H.
- Franklin S.S.,
- Gustin W. Iv..,
- Wong N.D.,
- et al.
- Schocken D.D.,
- Benjamin E.J.,
- Fonarow G.C.,
- et al.
- Go A.S.,
- Mozaffarian D.,
- Roger V.L.,
- et al.
- Armstrong A.C.,
- Ricketts E.P.,
- Cox C.,
- et al.